Conceptualizing the Gig Economy and Its Regulatory Problems

Conceptualizing the Gig Economy and its Regulatory Problems

Nikos Koutsimpogiorgos, Jaap van Slageren, Andrea Herrmann, Koen Frenken

Copernicus Institute of Sustainable Development, Utrecht University

Corresponding author: [email protected]

This version: December 30, 2019

1. Introduction

The advent of online platforms has been considered one of the most significant economic changes during the last decade (Kenney & Zysman, 2016; Van Dijck et al., 2018). In the context of labour markets, online platforms are used to match the supply and demand of flexible labour. The emergence of such platforms fits in a longer trend towards increasing contingent work, labour market flexibility and outsourcing work to independent contractors

(Stanford, 2017; Estlund, 2018; Hyman, 2018).

Online platforms mediating flexible labour are generally headed under the term ‘gig economy’ (De Stefano, 2015; Frenken & Schor, 2017). The best known of these gig economy firms is Uber – the media’s ‘poster boy’ for everything deemed good or bad about work via online platforms. However, the rise of online platforms as intermediaries of supply and demand of flexible labour is surely not limited to the taxi branch. Odd jobs (e.g. via

TaskRabbit), cleaning (e.g. via Helpling), care (e.g. via care.com), food delivery (e.g. via

Deliveroo), or programming and translating (e.g. via Upwork) are among the examples of services that are increasingly traded via online platforms.

1 There is a shared expectation that, even though small at present, the amount of labour hired through online platforms will continue to grow and is expected to account for a significant part of the economy in the near future (De Stefano, 2015). Given these expectations, scholars, unions, and policymakers alike take a great interest into the phenomenon of the gig economy. In their debates, we witness a proliferation of definitions and claims, which reflect the newness and complexity of the phenomenon at hand. A lack of agreed conceptualisation and analytical framework may hamper the accumulation of academic understanding of the gig economy, as well as the political deliberation processes regarding regulation of the gig economy.

To offer an analytical framework for the rapidly increasing number of concepts and policy proposals, we identify four dimensions along which the gig economy has been distinguished from other parts of the economy. These dimensions include: (i) Online platform versus offline intermediation, (ii) Independent contractor versus employee, (iii) Paid versus unpaid, and, (iv) Service versus goods. Taking the lowest common denominator of these four dimensions as a baseline, one can define the gig economy as the ensemble of ex ante specified, paid tasks carried out by independent contractors mediated by online platforms.

Our framework also allows to take a broader definition of the gig economy, including a wider range of economic activities along each of the four dimensions, namely: intermediation by offline platforms next to online platforms, employees next to independent contractors, unpaid tasks next to paid tasks, and goods rented out in the ‘sharing economy’ next to tasks carried out in the gig economy.

Our four-dimensional framework does not only aim to clarify the fuzzy conceptual boundaries of the gig economy, but also points to the four essential directions for regulatory responses to societal concerns raised by the advent of the gig economy. Accordingly, the four pillars of our conceptual framework also map onto four substantial regulatory questions

2 related to the gig economy, namely: (i) whether online platforms mediating the supply and demand for gigs should be regulated differently from offline intermediaries performing the same function, (ii) whether gig providers mediated by online platforms should be regulated differently from employees, (iii), whether paid gigs should be regulated differently from unpaid gigs, and (iv) whether providing gigs should be regulated differently from sharing goods. These regulatory issues are currently at the centre of the debates surrounding the gig economy. Accordingly, we argue that the future development of the gig economy is essentially contingent upon political choices regarding the four regulatory challenges that follow from our framework.

This article is structured as follows. Section 2, based on existing conceptual propositions of the literature, introduces the four dimensions that we use to conceptualize the gig economy. Section 3 discusses the regulatory questions related to the four dimensions of the gig economy. Section 4 concludes that the future development of the gig economy will be chiefly politically determined, depending on the regulatory positions taken on the analytical dimensions we propose.

2. Conceptualizing the gig economy along four dimensions

Despite the massive interest in the gig economy, a widely accepted definition is still lacking among academics, policy-makers and practitioners. Some scholars avoid a general definition, focusing on a specific platform (De Groen et al., 2016; Birgillito & Birgillito, 2018; Green et al., 2018; Hara et al., 2018), or specific sector instead (Cramer & Krueger, 2016). Others refer to the gig economy as “digital labour markets” without further defining it (De Stefano,

2015; Eichhorst et al., 2017; Burtch et al., 2018). And, when looking at scholars who provide clear-cut conceptualisations of what they regard as the gig economy, substantial differences remain (Stewart & Stanford, 2017; Kuhn & Maleki, 2017; Healy et al., 2017).

3 Definitions carry immediate empirical implications. As some define the gig economy more narrowly and others more broadly, the size estimates of the gig economy differ substantially. When looking at The Netherlands, as an example, one report estimates the size of the Dutch gig economy to amount to 0.4% of the working population (Ter Weel et al.,

2018), while another estimates its size to be 10.6% (Pesole et al., 2018) of the working population.

Despite little agreement on the conceptualisation of the gig economy, it is possible to distil four dimensions along which definitions of the gig economy diverge between authors.

As depicted in Figure 1, they include: (i) Online platform versus offline intermediation, (ii)

Independent contractor versus employee, (iii) Paid versus unpaid, and, (iv) Service versus goods. We elaborate on each of these four dimensions below.

Figure 1. Four characteristics of gig economy, in narrow and broader senses

Employee Unpaid

Independent Paid contractor

Online Services intermediation

Offline intermediation Goods

4 2.1 Online platform versus offline intermediation

Most scholars see intermediation by online platforms, be it through an app or a website, as a key defining feature of the gig economy (e.g. Aguinis & Lawal, 2013; Stewart & Stanford,

2017; Wood et al., 2019). In this view, the advent of such online platforms has led to the advent of the gig economy. This view is also shared in policy reports of single countries, such as the United States (BLS, 2017), the United Kingdom (CIPD, 2017; Department of BEIS,

2018), Finland (Statistics Finland, 2017), Sweden (SOU, 2017), The Netherlands (Ter Weel et al., 2018) as well as for Europe as a whole (Pesole et al., 2018).

The logic of considering only platform-mediated work as belonging to the gig economy is based on two principal arguments. First, scholars that see online platforms as a defining feature of the gig economy tend to argue that the role of rating systems and algorithmic management fundamentally differentiates online platform intermediation from older forms of offline intermediation (temp agencies, telephone operators, offline bulletin boards, etc.) (Di Stefano, 2015; Shapiro, 2018; Duggan et al., 2019; Wood et al., 2019).

Second, online platforms did not only change the technology used to mediate supply and demand, but also the legal nature of relationships, replacing bilateral relationships by tri- lateral relationships involving a worker, a requester and the platform (Aloisi, 2015; Di

Stefano, 2015; Duggan et al., 2019).

Other scholars, however, do not consider online platform intermediation as a defining characteristic when conceptualizing the gig economy (Friedman, 2014; Kuhn, 2016;

Stanford, 2017). Instead, they understand the gig economy as a more encompassing phenomenon, including all flexible work arrangements of independent contractors – regardless of platform intermediation. Proponents of this broader conceptualisation are often economists, who argue that the platform in itself does not fundamentally change the nature of

5 the gigs that are carried out as an ex ante specified, paid task (the taxi drive, the cleaning job, the programming task, etc.). The economic effect of online platform mediation has been to lower transaction costs in the market for gigs, which does not necessarily mean that gigs mediated by online platform should be conceptualized as a separate economic activity than those gigs that are not mediated by online platforms.

2.2 Independent contractor versus employee

The second dimension on which definitions of the gig economy diverge is the nature of employment. Most studies emphasise that the supply of labour in the gig economy concerns

“individuals”, “taskers”, ”“freelancers”, “self-employed”, “independent workers”, or

“independent contractors” rather than employees (Friedman, 2014; Prassl & Risak, 2015;

Kuhn & Maleki, 2017; Meijerink & Keegan, 2019). This ‘freelancing’ aspect of the gig economy also entails that work is organised into specific tasks upon which gig workers and requesters agree upon ex ante, i.e. before the completion of the task. Ex ante defined tasks are typically, but not necessarily, carried out by independent contractors rather than employees.

The possibility to carry out gigs as independent contractor or as employees leads Prassl and

Risak (2015) to distinguish between internal and external gig work (or what they call

‘crowdwork’). In this context, internal work refers to gigs carried out by a company’s internal workforce and external work refers to gigs carried out by workers active on an online platform.

Those who consider only independent contractors to be part of the gig economy ignore the simple empirical fact that some online platforms, like Deliveroo, started off by employing their riders, and only switched to contracting independent contractors later on

(Zekic, 2019). Other platforms, such as Hilfr in Denmark, pioneered a hybrid model in 2019 where workers start with a status of an independent contractor, but can opt for an employee status after 100 hours of work (Aloisi, 2019). And in Germany, platforms for delivery

6 services such as Lieferando, offer – highly flexible – employment contracts where riders are paid by the hour (including their waiting time).

A question at the centre of contemporary legal debates is whether gig workers are to be considered independent contractors or employees (Aloisi, 2015; Di Stefano, 2015; Prassl

& Risak, 2015; Taylor et al., 2017; Prassl, 2018). The legal status of an “independent contractor” would imply a certain amount of autonomy, which may be questioned. While some platforms only act as a simple bulletin board for gigs, other platforms are more actively involved in the transaction of a gig (including matching, contracting, and pricing) as well as the evaluation of a gig (through timing, ratings, and reviews), which may in turn be fed back into the matching algorithm. The control that such platforms exert over workers casts doubt about the autonomy of workers and has in several court cases offered legal grounds for a re- classification of independent contractors into employees (Di Stefano, 2015; Loffredo & Tufo,

2018; Prassl, 2018).

2.3 Paid versus unpaid

In accounts of the gig economy, most scholars explicitly focus on paid work (Di Stefano,

2015; Taylor et al., 2017; Kuhn & Galloway, 2019). Hence, their notion of the gig economy refers to market transactions only, and can thus be measured, for example, through bank transaction data (Farrell & Greig, 2016; Farrell et al., 2018). The focus on paid work is understandable, in particular, from a scholarly perspective: while many investigate whether independent contractors mediated by online platforms should be considered employees (for which payment is a necessary condition) (Aloisi, 2015; Di Stefano, 2015; Healy et al., 2017), other focus on questions related to financial matters, such as the minimum wage (Stanford,

2017) or tax issues (Thomas, 2017).

However, the focus on paid work as a defining criterion of the gig economy also raises questions. First, there is a substantial component of unpaid work associated with paid

7 work in the gig economy. For example, waiting time for chauffeurs and couriers is not compensated if they have the status of an independent contractor. And, especially on platforms that organize digital services that are done remotely (such as data entry, programming, translation, etc.), gig workers spend a lot of unpaid time searching for gigs

(Berg, 2016; Wood et al., 2019).

Regarding voluntary work, we also witness the advent of online platforms matching supply and demand of voluntary work. Juxtaposing ordinary paid work to voluntary unpaid work in this context is not a trivial matter. Platforms may frame the work that they mediate as voluntary, but nevertheless suggest a financial compensation by the requester. And, in some instances such as Helpper in Belgium, the work is advertised with hourly pay rates, albeit with a pay below minimum wage.

2.4 Services versus goods

The final conceptual issue concerns the question whether the gig economy only includes individuals carrying out gigs by selling their own labour, or whether it should also include individuals who rent out their own assets. Most authors agree that the gig economy should be restricted to labour transactions, so as to differentiate labour platforms from capital platforms, where labour platforms refer to intermediation of ex ante specified tasks in the gig economy and capital platforms refer to individuals who rent out their own consumer goods in what is known as the sharing economy (Farrell & Greig, 2016; Frenken & Schor, 2017; Duggan et al., 2019).

However, this seemingly clear-cut conceptual distinction is often not fully applicable, because sharing assets also involves some amount of labour (Frenken et al., 2019). For example, the tenant of an accommodation via Airbnb also pays for reception and cleaning, which can be considered gigs (regardless of whether the home owner or someone else carries out these tasks). In this sense, renting out an asset to a consumer can also be considered an ex

8 ante specified task, just as a gig, albeit a rather capital-intensive task. Following this view, some scholars place labour (gig) platforms and renting (sharing) platforms under the same conceptual umbrella (Schor, 2016; Healy et al., 2017), and some policy reports also include the sharing of assets in their analysis of the gig economy (CIPD, 2017; Pesole et al., 2018).

2.5 Narrow and broader definitions

The four dimensions we identified in relation to gig economy and its conceptualizations span a four-dimensional analytical framework. Following this framework, the lowest common denominator can serve as the narrowest baseline definition of the gig economy, as ex ante specified, paid tasks carried out by independent contractors mediated by online platforms. It also follows from our framework that broader definitions of the gig economy are conceivable, including intermediation by offline platforms next to online platforms, employees next to independent contractors, unpaid tasks next to paid tasks, and goods rented out in the sharing economy next to tasks carried out in the gig economy.

3. Regulatory classification

Our discussion of the four dimensions of the gig economy makes clear that the concept of gig economy has fuzzy boundaries. The proliferation of definitions both in academia and in policy documents can thus be understood as a manifestation of the difficulty to draw sharp boundaries along each of the four dimensions. Our framework, then, is helpful to unravel the sources of conceptual divergences.

Furthermore, we use the four-dimensional framework to shed light on current debates regarding the institutionalisation of the gig economy. These debates are centred on the distinctions between online platforms and offline intermediation, between independent contractors and employees, between paid and unpaid work, and between services and goods.

The exact lines between these categories can be drawn differently in different countries and economic sectors. Consequently, just as we witness a plurality of conceptualizations of the

9 gig economy, we also witness a plurality of institutionalisation processes of the gig economy

(Thelen, 2018; Uzunca et al., 2018).

More specifically, the four dimensions we distilled from the conceptual debate surrounding the gig economy, constitute an analytical scheme that allows to systematically reflect on four current debates about regulatory classifications. These issues concern the questions: (i) How an online gig platform should be classified, (ii) How a gig worker should be classified, (iii) How to deal with unpaid and underpaid gigs, and, (iv) How to deal with rental services based on personal assets. We will discuss these four regulatory debates one by one, including their interdependencies.

3.1 Online platform versus offline intermediation

One of the major differences between online platforms and older forms of intermediation in traditional labour markets consists in the radically new way that the intermediation is performed, namely through the use of algorithms, reviews, GPS, and electronic payment systems. It is the novel way in which online platforms match supply and demand that have raised platform-specific regulatory issues, including algorithmic discrimination, privacy and the lack of transparency (Helberger et al., 2018; Van Dijck et al., 2018). These concerns lead to a range of new regulatory challenges, not just for gig economy platforms, but also for online platforms more generally (including second-hand market places, search engines and social media), which are beyond the scope of the current paper.

In the context of the gig economy, the specific issue at hand is the classification of an online platform, which certainly is a non-trivial question. Gig economy platforms generally present themselves as online intermediation services or “technology companies”. Under prevailing e-commerce law in Europe and the United States, such platforms cannot be held liable for the actions of their gig workers (except in very specific circumstances) (Cauffman

10 & Smits, 2016; Helberger et al., 2018). This has also been the starting point of the European

Commission in its reflections on online labour platforms (European Commission, 2016).

However, sectoral regulations may apply to online platforms to the extent that they perform similar intermediation functions as “offline platforms”. No example could better demonstrate this conundrum than the case of Uber and the way it has been regulated across the two sides of the Atlantic (Thelen, 2018). Uber’s launch in the United States was characterised by an aggressive marketing campaign to rapidly increase its network and legitimise its operations. At that point, Uber was in direct conflict with the established regulatory systems for taxi services in many US cities, which operated on the basis of a fixed number of licenses (medallions). Instead of backing down in response to the regulatory backlash, Uber branded itself as an agent of “positive disruption” in a monopolistic market and used its increasing user base as a tool to advocate and promote its business model to policymakers. The success of Uber (and that of similar platforms) led many politicians to adopt this narrative, and to develop a whole new regulatory category, branding them as

“network transportation companies” (Thelen, 2018). Europe, on the other hand, followed a different regulatory approach, rejecting Uber’s claim of “positive disruption”, while forcing the company to adapt its model to the existing regulatory framework. Although the process evolved somewhat differently between countries (Thelen, 2018; Uzunca et al., 2018; Pelzer et al., 2019), the common trend across Europe has been that Uber phased out its UberPOP service with unlicensed drivers and moved instead to a licence-based-only model across

Europe, accepting on their platform only drivers with a taxi permit. This trend was reinforced in December 2017, when the European Court of Justice ruled against Uber by classifying it as a transportation company, which further settled the debate at the European level (Curia,

2017).

11 The above example is telling with regard to how regulators choose to deal with the platform aspect of the gig economy. While in the case of the United States, the use of the platform was considered an innovation which effectively set Uber (and similar firms) apart from the taxi market resulting in a new, tailor-made regulatory framework, the very same innovation got classified as a transportation service in Europe. At the heart of this controversy is the question whether an online platform, as an innovation, creates a new market or whether it rather disrupts an existing one (Prassl, 2018). Advocates of the former view make a case for a kind of “technological exceptionalism” or “digital distinctiveness” of the gig economy, while those supporting the latter view question the true novelty of the online platform. The answer to the question of whether the gig economy should be regulated separately (as a platform business) or within the existing legal framework (developed for

“offline businesses”) is not self-evident. As a consequence, the regulatory response is not straightforward but rather contingent upon political choices and local contexts.

3.2 Independent contractor versus employee

Across Europe, the employment status of gig workers is probably the most central topic in the public debate of the gig economy (Aloisi, 2015; De Stefano, 2015; Florisson & Mandl, 2018).

As different employment statuses directly translate into different forms of social protection, working conditions, and representation of workers, the legal term used to describe gig workers in each country has direct effects on their rights and obligations. Furthermore, it has broader implications in the field of competition law and taxation (International Labour

Organisation, 2013).

The transformation of the working relationship from bilateral to trilateral relationships inescapably raises the question whether gig workers are to be classified as employees (Prassl

& Risak, 2015). Traditionally, work relationships have been bilateral, be it between a requester and an independent contractor or between an employer and an employee. In the

12 case of intermediation by platforms, however, this bilateral relationship develops into a trilateral work agreement between the work requester, the platform, and the gig worker. In the transaction process between the requester and the gig worker, both parties also establish a contract with the platform providing them with online services that the two parties use to realize that same transaction. This, in turn, blurs the limits between the traditional concept of employee and independent contractor (Prassl & Risak, 2015; Loffredo & Tufo, 2018; Duggan et al., 2019).

The EU Treaties fail to provide a uniform definition of what constitutes a “worker”, beyond the scope of the freedom of movement (De Stefano & Aloisi, 2018). Subsequently, the European Court of Justice developed an own definition of the concept of “employee”, which is also employed by the Commission to describe who qualifies as such within the field of the “collaborative economy” (European Commission, 2016). According to this definition, an employment relationship exists when “for a certain period of time a person performs a service for and under the direction of another person in return for which he receives remuneration” (Judgment of the Court, 1986, C-66/85, Deborah Lawrie-Blum v Land Baden-

Württemberg).

Importantly, this definition is structured around three main concepts: (1) a relationship of subordination, (2) the completion of an activity and (3) the remuneration of the activity completed. While gig workers mostly complete activities which are remunerated through monetary payment, the question of subordination is less clear-cut. On the one hand, workers are assessed by clients through ratings and reviews and monitored by platforms regarding their acceptance rates and speed of service. This information may be used by the platform to decide to cut off an ‘underperforming’ gig worker from the platform at any moment in time and without explanation. On the other hand, gig workers enjoy the freedom of whether, or not, to accept a gig request and remain – in most cases – autonomous with regard to what to

13 charge and how they carry out the requested gig. Thus, depending on a number of factors

(such as the use of ratings in ways that can be detrimental to the gig workers, or whether the price is set by the platform or freely agreed), the gig worker may classify as an independent contractor or may be entitled to the legal rights and obligations of a traditional employee (De

Stefano & Aloisi, 2018).

A related question that is less often posed in the context of the gig economy is how to classify someone as an independent contractor (Frenken et al., 2018). Importantly, the category of independent contractors is not simply a residual category for those who do not meet the classification criteria of an employee. Instead, the question to be answered is whether an independent contractor can exercise the freedom as an independent business. One constraint posed by many platforms is that gig workers can only hold one account, and receive one assignment at the time, meaning that gig workers are technologically restricted to grow their business by re-selling their assignments. The difficulties to classify a gig workers as an employee on the one hand and as an independent contractor on the other, creates a legal

“grey area”. Here, workers find themselves to be economically dependent on the transacting platform, while not benefitting from the employee status. At the same time, they bear all the risks of being an independent contractor but do not enjoy the same economic freedom that regular businesses enjoy (De Moortel & Vanroelen, 2017).

The unclear status of a gig worker leads to a situation where it is ultimately up to national courts to decide whether a gig worker performing platform-mediated work is to be understood as an “employee” or “self-employed”. As De Stefano & Aloisi (2018, p. 53) point out in the example of food delivery workers, “a courier performing the same activity can be classified as a quasi-subordinate worker in Italy, as a self-employed worker in France, as an employee in Germany, as a “zero-hours” contract worker in the UK, or as an intermittent worker in Belgium”. Logically, the task of defining who is an independent contractor and

14 who is an employee falls upon the judiciary, which has to apply existing laws on new cases.

This may, however, not generate clarity per se, even within a single country, because the same court may reach almost opposite conclusions on different but related cases, as the case of Deliveroo in The Netherlands exemplifies (Zekic, 2019). Originally, Deliveroo started by employing its riders, but decided in January 2018 not to renew the fixed-term labour contracts but to continue its operations with independent contractors as riders. One of the riders, with the support of the largest Dutch trade union FNV, sued the platform claiming that there was no fundamental change in the employment relationship between the two parties and that the collective labour agreement of the professional goods transport sector should continue to apply. The Subdistrict Court of Amsterdam ruled against the worker while recognizing the shortcomings of current Employment Law with regard to the gig economy and calling upon legislators to take action. The union FNV then asked the court to rule on the practice of Deliveroo as a whole, instead of the individual case. This time, the same

Subdistrict Court of Amsterdam ruled in favour of FNV, forcing the company to abolish its model based on independent contractors. The case is still ongoing as Deliveroo filed an appeal.

Given the regulatory complexity regarding the classification of gig workers, most stakeholders agree that their work status should be clarified. In essence, this is a regulatory and, thus, political question, because the classification of gig work – possibly differentiated by sector – has direct consequences for wage setting, social security and consumer welfare.

Four different regulatory solutions have been proposed as a way to convert a minimum set of rights to gig workers.

The first solution, mainly advocated by the unions, is to consider gig workers as employees, based on the control that a platform exercises over its gig workers (Aloisi, 2015;

Di Stefano, 2015). Existing law and regulations would simply continue to apply and benefits

15 accruing from the employee status would ensure the social protection of gig workers. The obvious implication of such a pathway would be that most platforms could not continue to operate their current business models. Instead, they would have to assume the role of employers, requiring the introduction of fixed working hours and pay while workers wait for gigs. It would however not imply that the services offered through platforms would cease to exist. Most probably, however, such services would become more expensive leading the gig economy to shrink in size, including the associated consumer surplus.

The second way to deal with the legal uncertainty surrounding the classification of gig workers is to introduce a third category next to those of employees and independent contractors (Healey et al., 2017; Prassl & Risak, 2017). The aim would be to grant gig workers access to a set of rights which they would not enjoy as independent contractor.

Importantly, though, such an expansion of the legal codex would run counter to the established legal practice of dealing with new phenomena within the realm of existing codices. Furthermore, some scholars argue that a new category of gig workers could result in increased labour-market segmentation and social inequality (Florisson & Mandl, 2018).

Intermediate categories already exist in some EU countries, notably, the ‘worker category’ in the United Kingdom. A well-known case where the intermediate worker category has been extended in order to incorporate gig workers was that of Aslam v. Uber

BV in London. Two Uber drivers turned against the company claiming that they are not independent contractors as maintained by Uber’s terms of service, but instead should be reclassified as “workers” within the scope of the existing labour law, which means that they are eligible for minimum wage, sick leave and paid holiday provisions. As De Stefano and

Aloisi explain (2018, p. 48), the judgement to extend the worker category to Uber drivers showed that the court denied: “the fact that the company exercises a mere enabling activity between two opposite groups of users”. In doing so, “the British court emphasises that Uber

16 does not provide the opportunity for individually negotiating the content of the obligation, while tasks are performed personally, with no possibility of being be replaced temporarily”.

One country, France, took the initiative to create a new category as a response to the rise of gig economy platforms, thereby extending the French employment law to incorporate gig workers (French Labour Law n.2016-1088), so as to bestow gig workers with a set of employee rights. These new provisions apply in all those cases where the platform exercises a high degree of control over the worker, as defined by the law. When recognised as such, the gig worker is entitled to protection from work accidents and work-related disease and enjoys the right to unionisation and collective action (Donini et al., 2017). Regarding other European countries, Risak and Dullinger (2018) mention the ‘employee-like person’ in Austria and

Germany and the ‘para-subordinate’ in Italy as examples of already existing intermediate categories that may become applied to certain gig workers in the future.

Creating a completely new category remains a politically risky endeavour with the possibility of far-reaching and unintended consequences. With the establishment of a third category, employees may lose rights if their employment status is downgraded to the third category (Cherry & Aloisi, 2018). This may explain the reluctance of policymakers to adopt such an approach, especially in contexts where most flexible labour has the legal status of an employee at a temp agency (as for example in Belgium and The Netherlands).

A third route is to reconceptualise the notion of employer altogether. This approach means to move away from an inelastic definition of the employment relationship, where five conditions need to be met in order to qualify a work relationship as an employer-employee relationship, namely (Prassl & Risak, 2017): 1. the inception and termination of the employment relationship, 2. receiving labour and its fruits, 3. providing work and pay, 4. controlling all production factors of production, and 5. undertaking an enterprise with potential profit and loss. A ‘functional’ conceptualization of the employer (Prassl, 2015),

17 instead, is one “in which the contractual identification of the employer is replaced by an emphasis on the exercise of each function — be it by a single entity (…) or in situations where different functions may be exercised from more than one locus of control” (Prassl &

Risak, 2017, p. 281). Following this functional concept of the employer, the latter can be a single entity or combination of entities (e.g. a combination of the requester, the platform, and the gig workers). What matters is who plays a decisive role in the exercise of a particular employing function, and who can then be regulated as such according to prevailing employment law. Hence, a functional approach could be a way to deal with the complexities arising from trilateral work relationships inherent to gig work mediated by platforms.

The incorporation of gig workers into collective labour agreements constitutes a final way of ensuring gig worker protection to some degree. Several unions have taken up this role, as it reinforces their role as social partners and could increase their membership base

(Donini et al., 2017; Johnston & Land-Kazlauskas, 2018; Lenaerts et al., 2018). The most telling example of this fourth approach towards gig worker classification comes from a country with wide union coverage and an institutionalised social dialogue: Denmark. In 2018 the service-sector union 3F signed a collective agreement with the platform Hilfr, active in the care sector. Gig workers can decide to opt in to become an employee of the platform once they had worked 100 hours for Hilfr (enjoying a minimum wage, holiday pay, sick pay and a contribution to their pension savings), or can decide to opt out (Aloisi, 2019).

3.3 Paid versus unpaid

Most would agree that the gig economy concerns economic transactions only, thus dealing with paid assignments rather than unpaid assignment associated with voluntary work and hobby activities. There are many examples of platforms that mediate supply and demand of voluntary work and hobby activities (such as crowdsourcing platforms, open-source software platforms, Wikipedia, or websites of voluntary organisations). One could, however, argue

18 that not all of these are voluntary or hobby activities, as some do work for platforms in the hope that they will be selected for future paid assignments, or that they will otherwise generate revenues, for example, through the publicity they generated on a platform.

Users of online platforms also leave reviews and comments on a platform’s website, which could be regarded as voluntary work to the extent that these users add content with economic value, but without receiving any financial compensation. Taking this argument to its extreme, one could regard any user of a platform as a provider of unpaid work, because any recorded activity on a platform can be used by the platform as information, most notably, for advertising purposes (Fuchs & Sevignani, 2013; Zuboff, 2019). This issue becomes particularly acute once platforms extract economic value from the data that platform workers generate without being compensated for it (Van Dijck et al., 2018). This, in turn, leads to the, to date, open political question whether users ought to be financially compensated for the free

“digital labour” they perform, while being active on online platforms (Savona, 2019).

A related issue concerns the uncertainty of payments. Conducting work without remuneration is illegal in modern legal systems. Nevertheless, there are examples of workers completing assignments for an agreed price but without receiving the actual payment, because the requester is free to decide whether, or not, to pay once the assignment is completed. On MTurk, for example, the requester can deem the work submitted to be unsatisfactory and refuse payment as there is no mechanism for gig workers to challenge this decision. Much more common are questions arising from remuneration below the minimum wage (if one exists), facilitated by the status of independent contractors that platforms assign to gig workers. This practice, if left uncontrolled, could lead a race-to-the-bottom of labour standards and salaries. This concern is particularly acute in economic downturns (when labour is abundant in supply) and for global platforms mediating online gig work (i.e. gigs that can be performed online), so that gig workers can be hired from around the world

19 (International Labour Organization, 2018). In the absence of supra-national regulation and global unions, such global digital marketplaces disempower labour, and may lead to lower wages, decreasing labour security and labour standards alike (Freeman, 2006; Olney, 2013).

The main requesters of online gig work are large firms in Western countries. Hence, a regulatory pathway that may be promising in these contexts is one in which requesters commit to “decent commissioning”. For example, the IG Metall together with other unions set up a Code of Conduct in 2016, signed by eight internationally operating platforms, that includes a “fair payment” principle following the local wage standards of the requester. And, in 2017 an Ombuds office was established to enforce the Code of Conduct and resolve disputes between workers and signatory platforms (International Labour Organization, 2018).

The issue of low pay is especially pertinent for those who earn their full income in the gig economy. Schor and colleagues (2018) find that workers who use platforms only to supplement their income generally feel empowered and pick the best-paid gigs at convenient times, while workers who are dependent on platform for their full income generally feel disempowered, having to accept low paid gigs and less convenient working times. One way to counter low pay, is the instalment of a minimum tariff for independent contractors, as pioneered by The Netherlands Authority for Consumers and Markets in July 2019 (Authority for Consumers and Markets, 2019).

3.4 Service vs goods

Scholars generally differentiate the online labour platforms in the gig economy from capital platforms in the sharing economy, where individuals who rent out their own consumer goods such as cars and houses (Farrell & Greig, 2016; Frenken & Schor, 2017; Duggan et al.,

2019). The distinction between services and goods is important in that earnings in the gig economy are generally considered as income and taxed accordingly, while earnings in the sharing economy may not be taxed at all (such as occasional second-hand sales, carpooling

20 and car-sharing), or otherwise tend to fall under specific tax regimes (such as earnings from home rental). One particularly subtle example that illustrates the importance of differentiating between services and goods in the gig economy is the distinction made between ride-hailing

(e.g. via Uber) and hitchhiking (e.g. via BlaBlaCar). While the former is generally regarded as work, and taxed accordingly as income, earnings from hitchhiking are generally considered as an untaxed remuneration for the gasoline cost incurred by the car owner, who shares an otherwise under-utilised asset, i.e. an empty seat (Frenken & Schor, 2017).

While the difference between labour platforms and capital platforms may be conceptually straightforward, the distinction is less clear-cut in practice. Most tasks that gig workers provide still involve the use of assets required to carry out the service (such as a computer, car, bike, drilling machine, etc.). Reversely, a consumer renting out her assets does not only extract rents from this asset, but also performs work by cleaning, maintaining and inspecting the asset upon return (or hiring labour to this end). Hence, both work and assets are involved as inputs in any service, even if one would intuitively make a distinction between gigs as completing a particular task and sharing as renting out a particular asset.

Online platforms, then, can be situated on a continuum, ranging from the mediation of highly labour-intensive gig work (e.g. cleaning, tutoring) to highly asset-intensive sharing services

(e.g. home-sharing, car-sharing), with some platforms being situated in between (e.g. ride- hailing, home restaurants) (Frenken et al., 2019).

Following this reasoning, the key difference between labour platforms and capital platforms (viz. gig economy and sharing economy) is not related to whether assets are involved in providing a particular service, but rather to the question whether an asset is used by a supplier of a service (in the execution of a task) or by a consumer (who rents an asset for personal consumption). Prices in the gig economy are based on the willingness to pay for a particular service in the form of an ex ante defined task. By contrast, prices paid in the

21 sharing economy are based on the willingness to pay for the asset being rented out, that is, the services that a consumer can extract from having temporary access to a particular asset as a consumer of that good.

Arguably, the main regulatory challenge related the question of sharing versus gig work is of a fiscal nature. Bringing earnings from performing gigs and from sharing assets under the same fiscal umbrella, would resolve the classification issue. However, this does not resolve the bigger problem of collecting taxes from earnings in the first place (Oei & Ring,

2017; Thomas, 2017). While taxes from employees are relatively easy to collect, because employers can be obliged to disclose their wage payments to the tax office, tax collection from gig workers and sharing consumers is much more difficult. In situations where payments are made via online platforms, current privacy laws make it difficult to oblige platform operators to disclose transaction data, which is especially true if platforms are operated from abroad. And, if taxes can be imposed automatically on transactions done via platforms in the future, those who want to avoid paying taxes may look for alternative platforms that only match supply and demand of labour, leaving the payment to gig workers and their clients.

Interesting, while the main approach in the United States is to classify gig workers as self-employed and tax them as such, the issue is far less clear in Europe, because of the diversity of legal classifications of labour between countries. The European Commission has made clear that individuals who “carry out independently economic activity […] through sharing economy platforms” fall within the scope of the Value Added Tax (VAT) directive

(Council Directive 2006/112/EC) and qualify as taxable persons (European Commission,

2015). Whether a gig worker is classified as an employee or as an independent contractor defines whether s/he will be considered a person subject to taxation (Pantazatou, 2018). The issue of independence is thus crucial as it constitutes the defining element of an activity being

22 subject to tax. If the platform is considered to be just an intermediary, the gig worker is obliged to collect and pay VAT tax. If the platform is considered to be an employer, the platform is subjected to the regulations of the VAT directive, while the gig worker has to pay regular income tax (Pantazatou, 2018).

For an important part, freelance work has always been informal, implying that workers did not necessarily declare their income. With the rise of online platforms, though, the sum of income that remains undeclared may increase substantially. For this reason,

Thomas (2017) suggests to simplify tax collection. For example, platform companies could withhold the taxes for their gig workers, but without being classified as employers. As a further simplification, Thomas (2017) suggests a “standard business deduction” for gig workers, which would take away the administrative burden they now face when keeping records and filling in tax forms. Such an approach would also make it possible to introduce different tax rates for gig workers on the one hand and assets sharers on the other. For example, income from gig work is exempted from tax in Belgium up until 6,000 Euro per year but income from home sharing is not (Frenken et al., 2019). This differentiation can be justified for redistributive purposes assuming that those who own expensive assets such as houses, planes and boats are high-income.

4. Summary and conclusions

The hiring of workers via flexible work arrangements, where the requester and worker are matched via an online platform, is an emerging form of labour transaction – often called the gig economy. Enthusiasts argue that the gig economy meets the wishes of both requesters and workers for more flexible work relationships, while sceptics worry about low pay and limited social security of gig workers. Although the gig economy is receiving wide attention, consensus on a concept of the gig economy is remarkably limited. The simplification of the

23 gig economy as “digital labour markets” refrains from a more elaborate explanation of what the gig economy actually is which, in turn, complicates empirical assessments of gig work.

In answer to these conceptual and empirical problems, we proposed to conceptualize the gig economy along four dimensions: (i) Online platform versus offline intermediation, (ii)

Independent contractor versus employee, (iii) Paid versus unpaid, and, (iv) Service versus goods. Taking the lowest common denominator of these four dimensions, one could then define the gig economy, in a narrow sense, as ex ante specified, paid tasks carried out by independent contractors mediated by online platforms. Importantly, our analytical framework also makes it possible to take a broader perspective by including offline intermediation, employees performing gigs, unpaid activities and the sharing of goods into the concept of the gig economy.

Furthermore, each of the four dimensions of our analytical framework point to one fundamental issue regarding regulatory classification, namely: (i) How a gig platform should be classified, (ii) How a gig worker should be classified, (iii) How to deal with unpaid and underpaid gigs, and, (iv) How to deal with rental services based on personal assets. In sum, the four-dimensional framework helps to understand not only the various facets of the gig economy but also the corresponding regulatory challenges.

Next to offering an analytical framework for understanding the conceptual and regulatory debates surrounding the gig economy, our framework can serve as a basis for future research. It could, for example, be applied to understand the differences in regulatory responses across countries (Thelen, 2018; Uzunca et al., 2018). Online platforms are very suited for comparative research designs as most platforms are active in multiple countries.

Similarly, our framework can also be used to study differences in regulatory responses across sectors. Indeed, as the exact functions and operations of platforms differ across sectors, regulatory debates and actions may unfold differently between sectors (Frenken et al., 2019).

24 Additionally, one can use the framework to understand the combined effects of regulatory options along each of the four dimensions. For example, classifying a platform as an employer would imply the classification of the gig worker as an employee which, in turn, would solve the problem of low pay and foregone tax. Classifying a platform as an electronic service, instead, would mean to classify the gig worker as an independent contractor, which would not solve the problem of under-payment unless collective bargaining or minimum tariffs are allowed under current competition law. And, classifying online labour platforms as temp agencies would possibly resolve the classification issue of platforms and workers as well. The existing regulatory regime for temp agencies – which still may vary across countries – could then be transposed to platforms (possibly with some adaptations). Platforms would then follow the collective wage agreement with temp agencies and facilitate tax collection by governments, and it would also differentiate gig platforms from asset sharing platforms. However, temp agencies have to comply with regulations that are currently incompatible with the independent-contractor model of most online labour platforms, where the workers decide themselves when to work and how to carry out a job. Hence, a re- classification of online labour platforms as temp agencies would also necessitate a re-design of the platforms matching algorithm and the associated business model.

Clearly, political choices along each of the four dimensions of the gig economy will have important implications for the future evolution of the gig economy and the ways in which platforms can be deployed to mediate online labour markets. If gig workers will become classified as employees and platforms as their employers, adjustments in the platform’s business models will follow, probably raising prices for customers. However, if regulation is more accommodating, so that gig workers keep their status as independent contractor while platforms are considered to be e-commerce entities, the gig economy will most likely continue to grow. In between these two extremes, one can think of applying a

25 functional definition of the employer to be more flexible on what grounds the employer status can be assigned (Prassl, 2015). Alternatively, ad-hoc sectoral regulations or collective agreements cam be established, depending on a specific assessment of labour conditions, consumer interests, or other relevant public values (Helberger et al., 2018).

Regulations of gig platforms may thus evolve into different directions depending on the national or sectoral context (Frenken et al., 2019). This case-by-case approach has also been advocated by the European Commission (2016). The resulting proliferation of regulatory regimes provides an opportunity to learn across contexts from the variety of regulatory solutions adopted and their economic and social effects. At the same time, the increasing regulatory complexity – faced by gig workers, clients and platforms alike – may frustrate the realization of potential benefits provided by online platforms and may also make it harder to agree on social security reforms that would protect independent contractors in a more comprehensive manner, regardless of whether they work via online platforms. Our aim has been to unravel this regulatory complexity along four dimensions, thus providing a multi- dimensional framework to assess regulatory reforms to come.

26 References

Authority for Consumers and Markets (2019). Independent contractors are allowed to make arrangements about a minimum income. The Netherlands Authority for Consumers and Markets (ACM), https://www.acm.nl/en/publications/acm-independent- contractors-are-allowed-make-arrangements-about-minimum-income (23 July).

Aloisi, A. (2015). Commoditized workers: Case study research on labor law issues arising from a set of on-demand/gig economy platforms. Comparative Labor Law & Policy Journal 37, pp. 653–689.

Aloisi, A. (2019). At the table, not on the menu: Non-standard workers and collective bargaining in the platform economy, EUIdeas, https://euideas.eui.eu/2019/06/25/

Aguinis, H., & Lawal, S. O. (2013). eLancing: A review and research agenda for bridging the science–practice gap. Human Resource Management Review 23(1), pp. 6–17.

Berg, J. (2016). Income security in the on-demand economy: Findings and policy lessons from a survey of crowdworkers. Comparative Labor Law and Policy Journal, 37, 543–576.

Birgillito, G., & Birgillito, M. (2018). Algorithms and ratings: tools to manage labour relations. Proposals to renegotiate labour conditions for platform drivers. Labour & & Law Issues, 4(2), 25-50.

BLS (Bureau of Labor Statistics) (2017). Labor Force Statistics from the Current Population Survey. U.S. Bureau of Labor Statistics.

Burtch, G., Carnahan, S., & Greenwood, B. N. (2018). Can you gig it? An empirical examination of the gig economy and entrepreneurial activity. Management Science, 64, pp. 5461–5959.

27 Cauffman, C., & Smits, J. (2016). The sharing economy and the law. Maastricht Journal of European and Comparative Law 23, pp. 903–907.

Cherry, M. A., & Aloisi, A. (2018). A critical examination of the third employment category for on-demand work. In: N. M. Davidson, M. Finck & J. J. Infranca (eds.), Cambridge Handbook of Law and Regulation of the Sharing Economy (pp. 316–327). Cambridge: Cambridge University Press.

CIPD (Chartered Institute of Personnel and Development) (2017). To Gig or not to Gig? Stories from the modern economy, survey report, London.

Cramer, J., & Krueger, A. B. (2016). Disruptive change in the taxi business: The case of Uber. American Economic Review 106(5), pp. 177–182.

Curia (2017). Judgment in Case C-434/15, Asociación Profesional Elite Taxi v Uber Systems Spain, SL., Court of Justice of The European Union.

De Groen, W., Maselli, I., & Fabo, B. (2016). The Digital Market for Local Services: A one- night stand for workers? An example from the on-demand economy. CEPS Special Report 133.

De Moortel, D. & Vanroelen, C. (2017). Classifying self-employment and creating an empirical typology. Dublin: Eurofound.

Department of BEIS (2018). The Characteristics of Those in the Gig Economy. Research Paper: 2018 no. 2. Department for Business, Energy and Industrial Strategy (BEIS), UK Government.

De Stefano, V. (2015). The rise of the just-in-time workforce: On-demand work, crowdwork, and labor protection in the gig economy. Comparative Labour Law & Policy Journal 37, pp. 471–504.

28 De Stefano, V., & Aloisi, A. (2018). European Legal Framework for Digital Labour Platforms. Publications Office of the European Union https://ec.europa.eu/jrc/en/publication/european-legal-framework-digital-labour- platforms

Duggan, J., Sherman, U., Carbery, R., & McDonnell, A. (2019). Algorithmic management and app-work in the gig economy: A research agenda for employment relations and HRM. Human Resource Management Journal, in press, https://doi.org/10.1111/1748- 8583.12258

Donini, A., Forlivesi, M., Rota, A., & Tullini, P. (2017). Towards collective protections for crowdworkers: Italy, Spain and France in the EU context. Transfer: European Review of Labour and Research 23(2), 207–223.

Eichhorst, W., Hinte, H., Rinne, U., & Tobsch, V. (2017). How big is the gig? Assessing the preliminary evidence on the effects of digitalization on the labor market. Management Revue 28(3), pp. 298–318.

Estlund, C. (2018). What should we do after work? Automation and employment law. The Yale Law Journal 128(2), pp. 256–326.

Eurofound (2019). Platform Work: Maximising the potential while safeguarding standards? Publications Office of the European Union, Luxembourg.

European Commission (2015). Working paper No 878, VAT Committee – Question Concerning the application of EU VAT Provisions. DG Taxations and Customs Union, Indirect Taxation and Tax administration. Available at: https://circabc.europa.eu/sd/a/878e0591-80c9-4c58-baf3-b9fda1094338/878%20- %20VAT%20treatment%20of%20sharing%20economy.pdf

European Commission. (2016). A European Agenda for the Collaborative Economy. COM(2016) 356 final, 16.

29 European Parliament (2017). Online Platforms: How to adapt regulatory framework to the digital age? Briefing, No 607.323.

Farrell, D., & Greig, F. (2016). Paychecks, paydays, and the online platform economy: Big data on income volatility. JP Morgan Chase Institute.

Farrell, D., Greig, F., & Hamoudi, A. (2018). The Online Platform Economy in 2018: Drivers, Workers, Sellers, and Lessors. JPMorgan Chase Institute.

Florisson, R., & Mandl, I. (2018). Platform Work: Types and implications for work and employment – Literature review. Dublin: Eurofound.

Freeman, R. (2006). The great doubling: The challenge of the new global labor market. In: Ending Poverty In America: How to Restore the American Dream. New York: The New Press, Chapter 4.

Frenken, K., Vaskelainen, T., Fünfschilling. L., & Piscicelli, L. (2018). An institutional logics perspective on the gig economy. SocArxiv, 26 October. Retrieved from: https://osf.io/ preprints/socarxiv/uqn9v/

Frenken, K., & Schor, J. (2017). Putting the sharing economy into perspective. Environmental Innovation and Societal Transitions 23, 3–10.

Frenken, K., van Waes, A. H. M., Pelzer, P., Smink, M. M., & van Est, R. (2019). Safeguarding public interests in the platform economy. Policy & Internet, in press, doi.org/10.1002/poi3.217.

Friedman, G. (2014). Workers without employers: shadow corporations and the rise of the gig economy. Review of Keynesian Economics 2(2), 171–188.

Fuchs, C., & Sevignani, S. (2013). What is digital labour? What is digital work? What’s their difference? And why do these questions matter for understanding social media? tripleC 11(2), pp. 237–293.

30 Green, D. D., McCann, J., Vu, T., Lopez, N., & Ouattara, S. (2018). Gig economy and the future of work: A Fiverr.com case study. Management and Economics Research Journal 4, pp. 281–288.

Hara, K., Adams, A., Milland, K., Savage, S., Callison-Burch, C., & Bigham, J. P. (2018). A Data-driven Analysis of workers' earnings on Amazon Mechanical Turk. In: Proceedings of the 2018 CHI Conference on Human Factors in Computing Systems (p. 449) (April). ACM.

Healy, J., Nicholson, D., & Pekarek, A. (2017). Should we take the gig economy seriously? Labour & Industry 27(3), pp. 232–248.

Helberger, N., Pierson, J., & Poell, T. (2018). Governing online platforms: From contested to cooperative responsibility. The Information Society 34, pp. 1–14.

Hyman, L. (2018). Temp: How American Work, American Business, and the American Dream Became Temporary (New York: Viking).

International Labour Organization (2013). Regulating the Employment Relationship in Europe: A guide to Recommendation No. 198. Geneva: ILO.

International Labour Organization (2018). Digital Labour Platforms and the Future of Work: Towards decent work in the online world. Geneva: ILO.

Johnston, H., & Land-Kazlauskas, C. (2018). Organizing On-Demand Representation, Voice, and Collective Bargaining in the Gig Economy (No. 994981993502676). Geneva: International Labour Organization.

Judgment of the Court (1983) 3 July 1986. - Deborah Lawrie-Blum v Land Baden- Württemberg. - Reference for a preliminary ruling: Bundesverwaltungsgericht - Germany. - Worder - Trainee teacher. - Case 66/85. Retrieved from: https://eur- lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A61985CJ0066, 06/12/2018.

31 Kenney, M., & Zysman, J. (2016). The rise of the platform economy. Issues in Science and Technology 32(3), pp. 61-69.

Kuhn, K. M. (2016). The rise of the “gig economy” and implications for understanding work and workers. Industrial and Organizational Psychology 9(1), pp. 157–162.

Kuhn, K. M., & Galloway, T. L. (2019). Expanding perspectives on gig work and gig workers. Journal of Managerial Psychology 34(4), pp. 186–191.

Kuhn, K. M., & Maleki, A. (2017). Micro-entrepreneurs, dependent contractors, and instaserfs: Understanding online labor platform workforces. Academy of Management Perspectives 31(3), pp. 183–200.

Lenaerts, K., Kilhoffer, Z., Akguc, M. (2018). Traditional and new forms of organisation and representation in the platform economy. Work Organisation, Labour and Globalisation 12(2), pp. 60–78.

Loffredo, A., & Tufo, M. (2018). Digital work in the transport sector: in search of the employer. Work Organisation, Labour and Globalisation 12(2), pp. 23–37.

Meijerink, J., & Keegan, A. (2019). Conceptualizing human resource management in the gig economy. Journal of Managerial Psychology 34(4), 214–232.

Oei, S., & Ring, D. M. (2017). The tax lives of uber drivers: Evidence from internet discussion forums. Columbia Journal of Tax Law 8(1), pp. 56–112.

Olney, W. W. (2013). A race to the bottom? Employment protection and foreign direct investment. Journal of International Economics 91(2), pp. 191–203.

Pantazatou, K. (2018). Taxation of the sharing economy in the European Union. In: N. M. Davidson, M. Finck & J. J. Infranca (eds.), Cambridge Handbook of Law and Regulation of the Sharing Economy (pp. 343–356). Cambridge: Cambridge University Press.

32 Pesole, A., Urzì Brancati, C., Fernández-Macías, E., Biagi, F., & González Vázquez, I. (2018). Platform workers in Europe: Evidence from the COLLEEM survey, Joint Research Centre, Publications Office of the European Union, Luxembourg.

Prassl, J. (2015). The Concept of the Employer. Oxford: Oxford Universtiy Press.

Prassl, J. (2018). Humans as a Service: The promise and perils of work in the gig economy. Oxford: Oxford University Press.

Prassl, J., & Risak, M. (2015). Uber, Taskrabbit, and Co.: Platforms as Employers- Rethinking the Legal Analysis of Crowdwork. Comparative Labour Law & Policy Journal 37, pp. 619–652.

Prassl, J., & Risak, M. (2017). The legal protection of crowdworkers: four avenues for workers’ rights in the virtual realm. In P. Meil & V. Kirov (eds.), Policy Implications of Virtual Work (pp. 273–295). Cham: Palgrave Macmillan.

Risak, M., & Dullinger, T. (2018). The Concept of ‘Worker’ in EU Law. Status Quo and Potential for Change. Brussels: European Trade Union Institute, report 140.

Savona, M. (2019). The Value of Data: Towards a Framework to Redistribute It. SPRU Working Paper Series 2019-21, University of Sussex Business School.

Schor, J. (2016). Debating the sharing economy. Journal of Self-Governance and Management Economics 4(3), pp. 7–22.

Schor, J. B., Cansoy, M., Charles, W., Ladegaard, I., & Wengronowitz, R. (2018) Dependence and precarity in the platform economy, unpublished paper, Boston College, 47 p.

Shapiro, A. (2018). Between autonomy and control: Strategies of arbitrage in the “on- demand” economy. New Media & Society 20(8), pp. 2954–2971.

33 SOU (Statens Offentliga Utredningar) (2017). Ett arbetsliv i förändring – hur påverkas ansvaret för arbetsmiljön? Stockholm.

Stanford, J. (2017). The resurgence of gig work: Historical and theoretical perspectives. The Economic and Labour Relations Review 28(3), pp. 382–401.

Statistics Finland (2017). Labour Force Survey 2017: Platform jobs, Helsinki.

Stewart, A., & Stanford, J. (2017). Regulating work in the gig economy: What are the options? The Economic and Labour Relations Review 28(3), pp. 420–437.

Taylor, M., Marsh, G., Nicol, D., & Broadbent, P. (2017). Good Work: The Taylor review of modern working practices. Department of Business, Energy & Industrial Strategy, UK.

Thelen, K. (2018). Regulating Uber: The politics of the platform economy in Europe and the United States. Perspectives on Politics 16(4), pp. 938–953.

Thomas, K. D. (2017). Taxing the gig economy. University of Pennsylvania Law Review 166, pp. 1415–1473.

Van Dijck, J., Poell, T., & De Waal, M. (2018). The Platform Society: Public values in a connective world. Oxford: Oxford University Press.

Weel, B. ter, van der Werff, S., Bennaars, H., Scholte, R., Fijnje, J., Westerveld, M., & Mertens, T. (2018). De Opkomst en Groei van de Kluseconomie in Nederland. SEO- rapport 2018-30 (The emergence an growth of the gig economy in The Netherlands, in Dutch).

Wood, A. J., Graham, M., Lehdonvirta, V., & Hjorth, I. (2019). Good gig, bad gig: autonomy and algorithmic control in the global gig economy. Work, Employment and Society 33(1), pp. 56–75.

34 Zekic, N. (2019) Contradictory court rulings on the status of Deliveroo workers in the Netherlands. Comparative Labor Law & Policy Journal, Dispatch, 1 July, 5 p.

Zuboff, S. (2019). The Age of Surveillance Capitalism: The fight for a human future at the new frontier of power. New York: Public Affairs.

35